This R03 application proposes to conduct secondary data analysis to develop statistical models that explain disparities in HIV/AIDS found among African American and white substance users. In the proposed study, we will employ advance statistical analyses (latent variable mixture models) to consider the separate and synergistic influence of neighborhood factors (drug arrest rates and poverty), substance use patterns and neurocognitive functioning on sexual risk partnerships and, in turn, HIV/AIDS. First, we will evaluate the possible specificity in relationships between distinct neighborhood level factors (segregation, drug markets, poverty) and sexual partnerships and HIV/AIDS. Aim 2 of the study is to model the influence of neighborhood characteristics, substance use patterns, incarceration and neurocognitve functioning on sexual risk partnerships and estimate prevalence and incidence over an18-month period. The third aim is to estimate the magnitude of the association between sexual partnerships and being HIV-infected. The possibility of effect modification is considered with a focus on race and gender. If successful, this project should improve prevention science by: (1) increasing understanding of sexual risk partnerships among drug users and how they vary across race and gender; (2) providing support for considering factors beyond traditional SES indicators when attempting to understand racial disparities in HIV by highlighting the importance of contextual influences of drugs (e.g., rate of drug arrests) on sexual partnerships; and (3) specifying potential neurocognitive processes that might mediate/moderate the relationship between neighborhood, substance use and sexual risk partnerships. This study is the first in a series of innovative studies designed to explain racial disparities and prevent HIV/AIDS among African Americans. [unreadable] [unreadable] PUBLIC HEALTH RELEVANCE: The intersection between substance use and HIV is a major concern as core and peripheral members of drug networks represent a significant portion of HIV/AIDS cases. Although all persons involved in drug networks are at increased risk for HIV, African Americans tend to be disproportionately affected. Because interventions focused solely on the individual have had limited success in slowing the spread of HIV among African Americans, it is important to elucidate the influence and model the pathways through which social and interpersonal factors influence HIV acquisition. [unreadable] [unreadable] [unreadable]